Methods: Root canal treatment was performed on 33 cadaver teeth in situ two weeks prior to root resection and ultrasonic retropreparation (S12/90°D-tip on Suni-Max, Satelec, France), after which the teeth were retrieved from the cadavers. Two teeth were kept as positive and negative controls. The other teeth were divided in three different groups at random, with each group receiving one of the retrofill materials. Retrofills were exposed to water 5 minutes after placement. The teeth were stored at 37°C for 12 hours after which the root canal filling was removed. Microleakage (L in ìl/day) was measured for 24 hours under a pressure of 1.2 atm using a fluid transport model (FTM) and recorded as L=0, 0 Results: At all times FIX leaked less than MTA and IRM, but no statistically significant differences could be demonstrated. On the whole, leakage diminished after one month to re-increase after 6 months. When comparing the 2 techniques, the CFP demonstrated through-pores in all teeth, whereas with the FTM in only 14 of the 31 teeth through-pores could be demonstrated. Conclusion: Under the conditions of this study, despite some variations, no statistically significant differences among the three materials could be demonstrated. Based on the present results it appears that with the CFP more through-pores can be demonstrated than with the FTM.